作者: Weishan Zhang , Pengcheng Duan
DOI: 10.1109/UIC-ATC-SCALCOM-CBDCOM-IOP.2015.194
关键词:
摘要: Predicting resource demands in cloud computing environment is very important order to make system run optimally. The existing work falls short conducting prediction an satisfiable accuracy. In this paper, we propose use Deep Belief Network(DBN)-based approach for demanding prediction, which can capture high variances metric data without hand-crafting specified features. We have evaluated the proposed with Google cluster trace released 2011 show effectiveness terms of It shows that DBN-based predict term a accurate way, and long acceptable